The embodiments described herein relate generally to image processing. More specifically, the embodiments described herein relate to classifying and grouping images based on detecting image characteristics.
Many people enjoy attending amusement parks, such as theme parks, water parks, etc. Patrons of these parks typically like to memorialize their experience by taking pictures of the sights and activities. One option for a patron is to take their own personal pictures around the park. However, taking personal pictures requires that the patron carry around a potentially expensive device that could get damaged or stolen. Furthermore, taking pictures with a personal device limits the types of pictures that can be captured. For example, taking pictures while on many types of rides is typically infeasible, if not expressly forbidden.
To address the desire for patrons to have photographs taken of their experiences while on a ride, many parks have installed camera systems designed to take pictures at one or more strategic points along the trajectories of various rides. For example, a camera may be situated around the bottom of a roller coaster hump to take a reaction shot of a patron screaming and shouting as she is plummeting toward the ground. These cameras are generally located a select rides, and the cameras that are present tend to be configured to only capture still images. The captured images are usually available for purchase at a gift shop or a kiosk located outside of the ride, and may either be printed on-site or sent to the patron via mail or e-mail. The images for sale tend to be rather expensive to purchase. At the same time, even if a patron is interested in purchasing a photograph as a souvenir, there are typically long lines or waits at the gift shop or kiosk. Accordingly, these negative externalities reduce the likelihood that a patron will be willing to purchase the photograph, which corresponds to a loss of revenue.